Stats again! (and blackaces13's rant...)

[Zetack]

Ah, pokertracker... Ok I know I have a small sample, so please don't give me the: small sample dude, post. The thing is I really enjoy poker, playing it, discussing it, thinking about it. So I like playing around with this program, its fun.

Since I got pokertracker, I feel like I've been running badly. Which is annoying since it makes my pokertracker stats kind of ugly. Of course, I might just suck. And maybe its just an impression, you know the old annectodal evidence problem. Regardless, over the past 7200 hands I am down about 23 BB's.

So I found this interesting stat. Under Misc Stats it breaks showdowns seen by type of final hand. And my one pairs are getting slaughtered. Of the one pairs I've gone to showdown with, I've won a hair over a third of the time and am down a whopping 67 BB's. 67!!! This seems to confirm that I'm running badly (as opposed, ya know, to just plain sucking) since, unlike blackaces, I do expect my pairs to hold up more often than not (what would be a good percentage I wonder?) And I certainly don't expect to be losing money on them.

A couple of notes. I tend not to push real hard with second or worse pairs and I can release pocket pairs when I think the board or the action warrants it. So usually when I reach showdown with a pair its a pretty solid pair.

What do other people's stats with one pair look like? Particularly anybody who is on a nice winning stretch? I am correct in thinking that one pair should be an overall winner for me and that the long run will turn this around aren't I?

--Zetack

[blackaces13]

Zetak,

Do you play a lot of .5/1 at Party? I think this has a lot to do with what I'm complaining about here. I really need to get pokertracker myself, then I can save time screaming and crying and getting people to break it down for me because I could simply look it up myself. However, if you do play a lot of micro then -67 BB is rather catastrophic for only one type of hand isn't it? Perhaps I really am onto something here, or maybe I just overplay TPTK after the flop, I find it very hard to fold and often overcall on the river with it.

Oh yeah one more thing, SMALL SAMPLE DUDE. /images/graemlins/grin.gif

[bisonbison]

I don't have my stats in front of me, but I'm at around 4BB/100 over the last 25k hands, and I think I'm a loser in the high-card and one-pair hands.

Remember that PT counts your hand as one pair if the board is paired, or if you're chasing a flush draw and you pair your kicker, or any number of other things.

I don't think that set of stats is particularly illuminating.

[MortalNuts]

hey zetack --

I don't have my PT stuff handy, but winning a third of your single-pair showdowns doesn't sound that terrible to me.

Is more than a third of the money in the pot at showdown typically money you put in? My guess is, probably not. Unless you're pushing hard in heads-up situations and losing them, I guess.

Obviously that's a simplification, since not all pairs are created equal. But whatever.

I'll let the people who can actually quote you their own numbers give you real advice, though. /images/graemlins/wink.gif

Oh, I will say one more thing: some of the aggressive plays with pairs that you're probably (correctly) doing strike me as pretty high-variance. E.g., raising a limper or two in LMP with 99 -- it's usually the right play, but it's pretty easy to have a long stretch of hands where you lose some money with the hand. (You don't improve, and say an overcard doesn't fall till the river. Or some guy with KT gets nervous when the flop comes Txx, because "your PF raise means JJ or better," and never tells you you're beat. I have actually heard that line at Stars, btw.)

so yeah, probably it's just that your sample size is small. /images/graemlins/wink.gif

cheers,

mn

[Zetack]

[ QUOTE ]
I don't have my stats in front of me, but I'm at around 4BB/100 over the last 25k hands, and I think I'm a loser in the high-card and one-pair hands.

[/ QUOTE ]

I'd be interested to know for sure about the one pairs when you get that data in front of you. I don't expect to have a great showing for just high card, btw.

[ QUOTE ]

Remember that PT counts your hand as one pair if the board is paired, or if you're chasing a flush draw and you pair your kicker, or any number of other things.

[/ QUOTE ]

Sure, I guess, though, we call a bet on the river with a busted flush draw and a paired kicker because we expect the kicker to hold up often enough to make that call worth it right? (i.e, more than once in ten, when there's ten bets in the pot). So over time we shouldn't lose money with that call. And I know that stat includes the times you have 9-3 off in the BB and it gets checked around to the river and somewhere in there you hit a three and after the final check around on the river somebody's paired five beats you. But it seems like it'd take an awful lot of those to drive this number significantly into the red.

--Zetack

[Zetack]

Nah, I play a lot of .5/1.00 and .25/.5 at pokerstars


[ QUOTE ]


Oh yeah one more thing, SMALL SAMPLE DUDE. /images/graemlins/grin.gif

[/ QUOTE ]

LOL.

--Zetack

[Festus22]

Why do you think that EVERYONE who tracks their game has A-A as the number 1 winner? Is it because they always seem to make a set, get lucky with a 4-flush board and they have the ace of that suit? I think not. It's because A-A holds up as the winning hand unimproved (most report a win rate around 70%. I don't have my stats handy but let's use that as a baseline). So when one looks at A-K for example, you'll flop TPTK around 1/3 of the time. And this hand should win around 70% of the time using the A-A baseline. So A-K should win around 23% of the time. If you're not making money with a hand that has that high a win rate, I think you need to examine your play.

[Zetack]

[ QUOTE ]
Why do you think that EVERYONE who tracks their game has A-A as the number 1 winner? Is it because they always seem to make a set, get lucky with a 4-flush board and they have the ace of that suit? I think not. It's because A-A holds up as the winning hand unimproved (most report a win rate around 70%. I don't have my stats handy but let's use that as a baseline). So when one looks at A-K for example, you'll flop TPTK around 1/3 of the time. And this hand should win around 70% of the time using the A-A baseline. So A-K should win around 23% of the time. If you're not making money with a hand that has that high a win rate, I think you need to examine your play.

[/ QUOTE ]

Hey Festus,

Thx for your feedback although this sounds more like its directed towards blackaces. My expectation is that pairs that reach showdown aren't a type of hand that you expect to take a beating on. Of course that stat encompases more than just top pair top kicker, so you would expect it to it do worse than TpTk, but do you expect it to do as badly as I've been doing over this admittedly small sample? So I'm interested in what other's stats look like, particularly if they've been playing better than my pretty close to break even poker for the last 7200.

Your post did spur me to look at A-K off and as you would expect it is a nice winner for me overall. On the other hand, perhaps not as nice as it could be. My impression that it and other solid pairs aren't holding up as nicely as I'd like seems reinforced when I notice how far down A-K is on my list of winners on a BB won per time I've had the hand basis. K-J suited, for example, has done far better for me than A-K off over this (limited) time frame. K-Qoff is an actual loser. Heck, J-7 suited is better on a BB/hand basis than A-K off.

I could go on, but this is really parsing things pretty fine with way too few samples. Which is why I simply looked at that down 67 BB's for pairs that see showdown and thought that looks like either a really bad stretch of play, or an unfortunate run of cards. And I thought I'd see what other people's stats looked like. Again, my expectation is that in the long run, if I ever hit it, and given decent play (assuming I'm capable of that), that stat should be nicely in the black (or green as pokerstracker would have it). Am I wrong about that?

--Zetack

[MaxPower]

The pokertracker stats can be very misleading depending on what site you play. If you play on more than one site these things are calculated differently for each site.

There are many scenarios in which pokertracker will tell you that you lost money with one pair, but it is misleading. Lets say you have a flush draw and the board pairs on the river or you pair one of your hole cards and it is checked down. Let's say, you hit bottom pair and it checked around on all streets and you lose to a higher pair. All these little scenarios add up and therefore the stat for one pair is very misleading.

As I said in my other response, when you are running bad your one pair hands will be beaten a lot. Looking at my pokertracker data, I have had a run of 10,000 hands where I lost 300BB with one pair. I ran very bad over those 10,000 hands and lost about 100BB. If a few more of those one pair hands had held up, my results would have been alot better.

[Zetack]

[ QUOTE ]


As I said in my other response, when you are running bad your one pair hands will be beaten a lot. Looking at my pokertracker data, I have had a run of 10,000 hands where I lost 300BB with one pair. I ran very bad over those 10,000 hands and lost about 100BB. If a few more of those one pair hands had held up, my results would have been alot better.



[/ QUOTE ]

Yes that is my assumption--I think the stat reflects that I'm running poorly right now, I'm trying to figure out if that's a correct assumption. So pick say a ten thousand hand section where you were running what you'd consider normally. What does that stat look like then?

--Zetack

[MaxPower]

By the way, go into pokertracker and use the replay feature to watch all the one pair hands. You will see what I am talking about.

[MaxPower]

[ QUOTE ]
[ QUOTE ]


As I said in my other response, when you are running bad your one pair hands will be beaten a lot. Looking at my pokertracker data, I have had a run of 10,000 hands where I lost 300BB with one pair. I ran very bad over those 10,000 hands and lost about 100BB. If a few more of those one pair hands had held up, my results would have been alot better.



[/ QUOTE ]

Yes that is my assumption--I think the stat reflects that I'm running poorly right now, I'm trying to figure out if that's a correct assumption. So pick say a ten thousand hand section where you were running what you'd consider normally. What does that stat look like then?

--Zetack

[/ QUOTE ]

It probably says that I am either a very small loser or a very small winner with one pair, but it is very misleading for the other reasons I stated in my post. I just thought of one more reason those stats are problematic. I often win one pair hands without a showdown and I assume that happens occasionaly in .5/1 games as well (I believe this is not included in this stat when you are playing on Party).

[blackaces13]

Interesting that you mention AA. Actually, I don't think it does win very often unimproved against super loose players. The thing is that there are a lot of Party .5-1 games that aren't THAT loose. Games where you actually can get it down to 1 or 2 opponents witb a UTG raise. Now I know you don't want this and I'm not suggesting that you want this what I'm suggesting is simply that it does happen which is why people may be able to boast a win % of 60 or higher with them.

However, most people know that when you play AA againt 9 random hands until showdown it wins about 30% of the time. Also, this 30% is far higher than any other hand and furthermore its incredible more than its "fair" share which would be 10% in a 10 handed game. Fine, but what is interesting to note, and a fact that no one ever seems to mention, is that those Aces will make a set 20% of the time.

So lets to a little extrapolation. 30% wins, 20% sets made = 10% wins with unimproved aces. Hmmmm, seems that they're only breaking even now huh? Actually, this isn't even enough info since the sets don't ALWAYS win. So how about we not include the times there is a 4 flush and you have the ace or the times you make a broadway in the other 10% and call it even, for the times the sets lose we'll chalk it up as equalled by the times you hit a straight or flush and maintain the 20% coming from "improved" hands. I'm not sure this is right but I think its at least close.

Hence, unimproved aces actually win ONLY their fair share against 9 random hands. Add to this the fact that they are near impossible to fold and you can see how my gripe with TPTK may actually hold a little bit of water. Something to think about anyway since you never hear anyone get into the unimproved aspect in much detail.

[Festus22]

As I mentioned, I don't have my results handy. I remembered sthief09 had just posted his results after 10K hands. His biggest winners were:

best hands by BB/hand
AA
KK
QQ
AJs
AKo
AKs
KJs
77
QJs
KQo
44

Clearly his 4th & 5th highest winners likely made TPTK and held up. If your overall single pair hands are deep in the red, perhaps you're calling down too much. It's tough to say when we're talking in generalities.

Something that helped me A LOT was emailing a 100 hand history block to Joe Tall and he added comments where he would do something different. I'd be happy to do the same for you if you think the advice I'd offer would be helpful. PM me if you're interested.

[eric5148]

I really don't think this stat means a whole lot for reasons bisonbison and maxpower have said. I've played 12,070 hands combined of .25/.50, .50/1, and 1/2. My total net is +$367, with one pair it's -$171. This is with 51% won $ at showdown for one pair.